Upload your homework document here
Each of the two images below contains a hidden feature (or signal) that encompasses 200 pixels or so. While you can try to recover this signal (which is weak which is why you can't see it) from the structured noise fields that make up the background any way you like, this will be difficult. It is better to do the sliding box pixel statistics method discussed in class on friday and this is part of the assignment anyway
Steps:
- Download the two files
Note - you can use this python code to produce the image:
import numpy as np
from matplotlib import pyplot as plt
row,col,data=np.loadtxt("noisyimage.txt",unpack=True)
rsize = int(max(row))
csize = int(max(col))
data=np.array(data).reshape(rsize,csize)
plt.imshow(data, interpolation='None',cmap=plt.cm.Greys_r)
- Include the gray scale depiction of each image in your write up
- Now slide boxes of different sizes over the array and calculate the mean and standard deviation in each box. Remember that too large of box will average over the structured noise and too small of box will make each of the positive and negative features of the structured noise have a significant signal.
For both Image A and B produce a table (might be large) that shows Box Size, Center Coordinates of Box, Mean pixel value in the box, and standard deviation of those pixel values
- Now make some kind of plot that summarizes the pixels statistics in the box as a function of box size. Submit that plot
- Now THINK about the results and make an argument for where and why you think the signal is present.
- With this new information, try to do some kind of other iteration with the data to make a mask that you can subtract from the original image to see if the feature can popout. Submit the subtracted imaged and an explantion of what you did - Note, I don't expect anyone to be successful at this.
|